That feeling maps directly to what researchers just found about how ChatGPT decides who to recommend. The game is being played before most people even know there's a game.
“Ok, but who’s going to lead it?” I asked. “You are” was the answer.
At Ticketmaster, we needed a team of people to begin moving our infrastructure to the cloud. We had good engineers and a software architect to work on it. But, we didn’t have a leader. In that meeting, it became clear that my peers and sponsoring executive had already chosen someone: me.
I didn’t even realize I was an option.
That feeling maps directly to what researchers just found about how ChatGPT decides who to recommend. The game is being played before most people even know there's a game.
Researchers found that when ChatGPT generates a recommendation, it often already has a preferred answer baked in before it runs any external search. Appearing inside that pre-search internal query is worth 33 times more than getting cited after the search runs. That number changes the entire priority stack for anyone doing generative engine optimization. If your brand is not in the model's internal knowledge, a citation in a blog post somewhere may not save you.
AI-referred traffic is growing fast, but it converts differently than search traffic, and most sites are not set up for it. Visitors arriving from ChatGPT or other LLMs tend to be further along in their thinking, more specific in what they want, and quicker to leave if the page does not match their expectation precisely. The article lays out concrete steps for adjusting landing pages, calls to action, and content structure to turn that traffic into actual revenue instead of just sessions in your analytics report.
Google is moving fast on AI Max for Search campaigns. Starting in September, Google will begin auto-migrating legacy Search campaign features to AI Max, whether advertisers are ready or not. For agency owners and PPC managers, this is not a future problem. Client briefs need to be updated now, and campaign structures that rely on older keyword controls may behave very differently once the migration runs.
The 33x finding from Search Engine Journal is the kind of number that should stop you mid-scroll. It means that if ChatGPT already thinks of your brand as an answer to a question, you are 33 times more likely to get recommended than if you only show up in a post-search citation. Think about what that means practically. A law firm that has been consistently publishing clear, authoritative content about estate planning for three years is probably already in the model's internal knowledge. A firm that just started a content push last month is not, no matter how many backlinks they build this quarter. The implication is that GEO is not a short game, and the brands that treated content as a long-term asset rather than a campaign are now sitting on something very valuable.
The conversion piece from Search Engine Land adds a layer that most teams are skipping. Say you do start showing up in AI recommendations and traffic climbs. If your homepage or a generic service page is where that traffic lands, you will lose most of it. AI-referred visitors already know roughly what they want. They are not browsing. They are checking. A dental practice that gets recommended by ChatGPT for "best Invisalign provider near me" needs a page that immediately confirms they are in the right place, shows proof, and makes the next step obvious. A generic "Welcome to our practice" page will not cut it, and the bounce will tell Google nothing useful either.
The Google AI Max timeline is the forcing function that ties everything together. Agencies that wait until October to figure out what AI Max means for their clients' campaigns will be explaining bad results instead of preventing them. The smarter move is to audit which campaigns rely on exact match or other legacy controls right now, flag the ones most at risk, and get ahead of the conversation with clients before Google makes the decision for you.
The 33x multiplier is real, but it is also a trap if it sends people chasing LLM citations the wrong way. We have seen course creators and agency owners pivot hard toward "get mentioned in AI answers" as a tactic, usually by stuffing their content with brand mentions and hoping for the best. That does not work. What actually builds internal model knowledge is the same thing that built Google authority ten years ago, which is consistent, specific, genuinely useful content published over time on a topic you actually own.
A solopreneur who has written 80 detailed posts about e-commerce email strategy for Shopify stores is far more likely to be in ChatGPT's pre-search knowledge on that topic than a larger agency that published a few broad "email marketing tips" roundups. The AI Max migration is a separate but equally urgent issue, and the agencies that treat it as a technical checkbox rather than a strategic shift are going to have a rough Q4 explaining to clients why performance changed.
What is this shift looking like in your own work? Bring what you are seeing to the IMG community and compare notes with other operators working the same problems.
Join the IMG Community →— Tim Nichols & The IMG Team
